ABSTRACT
Since the founding of the European Union, religion has become an increasingly important aspect in shaping European identity and thereby social cohesion in Europe. Social cohesion depends to a high degree on a successfully established distinction between ‘us’ and ‘them’. Religion became one important marker of such boundaries. However, religion works both as an individual-level and an EU-level property. In order to take religion into consideration on the level of individual religiousness and as an institutionalised framework, we combine an analysis of key documents of the European Commission with a quantitative analysis of individual attitudes towards Europe. This combination of methods enables us to track the ‘discovery of religion’ by the European Commission as a means for social cohesion and the potential of religion to create Europeanness among the EU-citizens. We focus on the crucial period between 1990 and 2000 in which the major system transformation from EEC to EU took place. The quantitative analysis is conducted as a multi-level analysis on the basis of the European Value Survey 1990 and 2000. The data reveal that in fact over a 10-year period, the European integration project has begun to evolve towards an identity project with religion as a key factor on both levels.
1. Introduction
Since its founding in 1957, integration has been the central goal of the European Union. What then seemed a necessity of the time to secure peaceful democracies after two world wars and to overcome economic shortages has now turned into a comprehensive social and political project that aims at amalgamating people of different origins and cultures. One of the politically pressing questions became as on what grounds such different societies and peoples can be integrated.
In the early 1990s, the European Commission (EC) discovered ‘religion’1 as a political arena and a resource for integration. Along with the transformation of the European Economic Community (EEC) into the European Union (EU) in 1992, not only social cohesion was discovered as a synonym for integration2 but also the role of ‘religion’ was strengthened as a means for social cohesion through legal documents. In the Commission's perspective, religion became a resource – amongst others – for social cohesion in Europe and consequently, for the integration of the European people.
While ‘religion’ was recognised on the political level as facilitating social cohesion, it remained an open question to what extend religion and religiousness serves feelings of integration on the individual level. Sociological theories give raise to contradictory expectations: while according to empirical research, religion enforces social cohesion; secularisation theses suggest that religion becomes less salient in modern societies and thereby, its effects on social cohesion looses significance.
In this article, we analyse whether European religion(s) and religiousness and their institutional regulation support European integration. We want to know (a) how the Commission uses religion as a means to social cohesion and (b) whether such means-end-combination of integration holds for the individual level – more precisely: does individual religiousness correlate with individual Europeanness?
Our historical analysis covers the critical period before and after the system transformation of the EEC into the EU (1990–2000). This period is important for understanding European integration developments: it covers the period in which social cohesion became a goal in its own right. In order to trace the dynamics on the EEC/EU-level and to analyse changes on the individual level, we employ a comprehensive concept of religion. We do not as much focus on how religiousness or its impact changed in Europe as on what its different dimensions means for European social cohesion. We thereby respond to shortcomings in the current debates on religion in Europe, which often ask either whether different dimensions of religion still matter (e.g., Davie 1990; Greeley 2002; Pollack 2008; Houston 2011) or what impact religion still has (e.g., Fukuyama 2001; Frejka and Westhoff 2008; Jagodzinski 2009), but not in which ways the different dimensions work. We also respond to shortcomings in the debates on EU-integration and Euro-scepticism neglecting religion as a factor altogether (e.g., McLaren 2002; Eichenberg and Dalton 2007; Weßels 2009).
The paper combines a document analysis of the European Commission's perspective3 on ‘religion’ with a quantitative multi-level analysis of the impact of individual religiousness on feelings of ‘Europeanness’. For the document analysis, we exclusively used legislative texts and policy papers published by the EC as a unique data source. For the quantitative analysis, we use the European Value Survey (EVS 1990 and 2000). Both methods in combination provide insights into the changes that took place in Europe during this period with regard to European integration. We by no means indicate that EEC/EU-institutions or developments on this level influence individual-level attitudes. It is, however, our aim to explore how religion became a means of integration on the EC's level and whether this proposed means-end-combination can be measured on the individual level in order to detect the different integration dynamics taking place. The analysis thereby contributes to a better understanding of the mechanisms of social integration by analysing the role of multi-level cultural factors such as religion.
2. Social integration in the European Community and European Union
The question of how cultural factors like religion shape social integration gains its relevance from three research traditions: political philosophy stresses integration and social cohesion as a necessity to uphold democratic systems, welfare-state research focuses on factors like class or income that support or hinder (European) integration, while sociology of religion discusses religion as a powerful belief system that still provides a common denominator for societies.
In this section, we discuss how social cohesion became the aim of the European integration project and its relationship to religion. We do so by tracing the political implementation of religion as a political means of social cohesion on the EEC/EU level and by summarising sociological perspectives on religion and its impact on social cohesion.
2.1. Integration by homogeneity
2.1.1. The EU perspective on social cohesion
EEC and EU have a long history in defining and politicising social cohesion, beginning with the Treaties of Rome (1957). The European integration process started with the idea that economic integration should develop spill-over effects to other policy fields. As argued repeatedly, this idea changed over time (e.g., Shore 2000; Dinan 2005; Wagener et al.2006; Münch 2008) and an increased interest in social cohesion in its own right developed in relation to the challenges of economic stagnation and political crisis in the late 1960s and the 1970s.
The conflict over decision-making procedures in the European Council in the late 1960s, the first enlargement in 1973 by Denmark, the United Kingdom and Ireland, and the first oil crisis (same year) caused feelings of stagnation and crisis among the EEC's political élites and the people of its member states (Dinan 2005: 39–69). During this crisis, the idea of economic integration and its spill-over effects lost its integrative dynamics. Against this background, the Commission empowered Leo Tindemans, former Prime Minister of Belgium, to envision how to revitalise European integration. He suggested developing social cohesion by creating common symbols e.g., a European currency, flag or anthem. In addition, Europe should develop its own social and political ‘dimension’ which would include social rights and policies and a political union (Tindemans 1976: D 53). In a longer run, this ought to create feelings of Europeanness – whereas Europeanness remained un-defined for the time being.
The 1970s crisis and the Tindemans Report served as starting points for a new direction in EEC policies. The Commission became the central institution of the newly framed integration process and started to strengthen social cohesion amongst the people of Europe by establishing different programs and by rearranging the European political arena with the Single European Act in 1986.4 Most important changes came with the EU, which in 1992 was established as a political union alongside the economic community. With cultural programs and by emphasising civil rights, the EU initiated policies ‘close to the people’. The Maastricht Treaty in 1992 marked the legal turning point in institutionalising social and political integration, facilitating integration far beyond economy (Treaty on European Union: Article A; BEPA 2010). The Treaty must be considered as ‘system transformation’: it changed the European project into an integrated economic, political and social enterprise (Kohler-Koch 1999; Jansen 2005; Eichenberg and Dalton 2007: 132) – a development that assisted the EU in enlarging its policy fields and in initiating an independent policy on culture and civil society (Shore 2000; Münch 2001; Quenzel 2005). At the same time, the EC began to incorporate civil society organisations like trade unions, churches and other interest groups into the decision-making process on the European level in order to advance the European integration project through different channels (Charter of Fundamental Social Rights, EC 1998; Official Journal 1999; The White Paper, EC 2001).
During this process, the Commission established its own notion of social cohesion5 and solidified it by several initiatives, e.g., ‘Agenda 2000: for a stronger and wider Union’ (EC 2000), the establishment of the Euroregions and the Structural Funds. The Commission defined social cohesion as the integration of the people of Europe beyond the nation-state by offering opportunities to identify with ‘Europe’. First Progress Report on Economic and Social Cohesion (EC 2002) stated:
European programmes need to focus on adding value beyond that which is possible at national level. (…) Such contributions speak (…) of an enhanced role for the Commission in supporting networking between regions, in promoting sustainable development and in the creation of a coherent overall vision as a frame of reference for cohesion policies(…). (similar: BEPA 2010)
The Commission emphasised the aspect of homogeneity of interests and beliefs as a basis for integration. The shift from mere economic integration to political and social integration established social cohesion as a goal in its own right: Europeanness was no longer considered as side-product of the Common Market but as the locomotive of a successful integration project.
2.1.2. The citizen's perspective on social cohesion
These developments give rise to the question as to how the European citizens perceived the integration project. Empirical studies on Euro-scepticism showed a strong relationship between the increasing economic benefits of and public support on the individual level for the European project until the 1990s. By 1991, citizen enthusiasm had reached a historic height. This support eroded in the post-Maastricht Treaty era (e.g., Eichenberg and Dalton 2007). While after the Single European Act of 1986, mostly economic costs and benefits seemed to be responsible for citizens’ attitudes toward the integration process (e.g., Gabel 1998), citizens started to judge the EU-integration with regard to the consequences perceived for national redistribution systems and cultural similarities among people after the Maastricht Treaty of 1992 (e.g., McLaren 2002). Although there is no doubt about the importance of economic factors, cultural factors as religion according to these results are important as well and their neglect in this debate so far is surprising (as an exception: Nelson et al.2001).
2.2. Religion facilitating European integration?
2.2.1. The Commission's perspective on religion
Although the European project started as a predominantly Catholic elites’ project (Nelson 2004; McCrea 2007; Spohn 2009), ‘religion’ was discovered by the Commission as a policy field and a resource for integration during the system transformation between 1992 and 2000. Our document analysis of the talk by the European Commission's President Jacques Delors in 1992 and its impact on other legal and policy documents reveal this development.
Delors addressed representatives of the Protestant and Roman Catholic Church and emphasised the churches’ ability to stimulate the integration process. With the notion of giving ‘a soul to Europe’, he expressed that the churches were capable of bringing people together and pledging them to the European project. He stated:
We are in fact at a crossroads in the history of European construction. 1992 is a turning point. (…) The Maastricht summit marked the end of the economic phase of European construction – what has been described as the ‘semi-automatic’ development of the EC, based on the drive towards Common Market. (…) – a time when the debate on the meaning of European construction becomes a major political factor. Believe me, we won't succeed with Europe solely on the basis of legal expertise or economic know-how. It is impossible to put the potential of Maastricht into practice without breath of air. If in the next ten years we haven't managed to give a Soul to Europe, to give spirituality, and meaning the game will be up. … This is why I want to revive the intellectual and spiritual debate on Europe. I invite the churches to participate actively in it. (…) We are working on the idea already. We must find a way of involving the churches. (quoted in Luibl 2005: 197)
According to this quote, churches offer religion as a spiritual and emotional resource. They are able to provide not only significance to the integration project but also a ‘soul’ or affective dimension (Moïsi 1999; Weiss 2003). By explicitly addressing the churches, Delors outlined a new policy field: Religion and Europe. Starting just with a handful of catchwords, this outline had consequences on the legislation level which substantiated and stabilised the new policy field.
In 1994, the Commission started the corresponding initiative ‘A Soul for Europe’ in order to incorporate the churches and, later on, other religious groups. This initiative offered a chance for dialogue between religious groups and the Commission which facilitated affecting each other's policies (GOPA 2010). The Commission's interpretation of the role of churches became two-fold: they not only add spirituality and emotionality – they are additionally seen as mass organisations that are able to mobilise support for integration: ‘Since these communities represent a large number of people in Europe, the actual number of members in ‘A Soul for Europe’ is much bigger than six. In fact, every citizen in Europe that belongs to any of these communities is, in a way, a member of the initiative’ (GOPA 2010). The following paragraph form the White Paper on European Governance (COM(2001) 428 final, 2001: p. 14) indicates the Commission's perspective on the role of religion and churches:
Churches and religious communities have a particular contribution to make. The organisations which make up civil society mobilise people and support, for instance, those suffering from exclusion or discrimination.
By enlarging its understanding of churches and religious communities, the Commission clarified its understanding of religion. In this quote, integration and the role of churches and religious communities became distinct: instead of addressing single individuals, the EC aimed at integrating individuals as part of religious communities. In the Commission's perspective, religious communities serve as intermediate organisations which should support the integration of people into the European project.6
Within the treaties, ‘religion’ was not mentioned until the Treaty of Amsterdam in 1997. In that document, the 11th Declaration contains the ‘Declaration on the status of churches and non-confessional organisations’. This declaration was the first legal outcome of the initiative ‘A Soul for Europe’. As part of the Appendix, however, the Declaration constituted a weak legal position. Only later, in the Treaty of Lisbon from 2008 (Art 16.1), this Declaration became an integrated part of the community law: ‘The European Union respects and does not prejudice the status under national law of churches and religious associations or communities in the Member States. The European Union equally respects the status of philosophical and non-confessional organizations’.
Until the ‘Church Declaration’ of Amsterdam, the EU had been ‘blind to the churches’ (Mückl 2005); religious communities now became explicitly recognised as actors on the EU level (Jansen 2000; Robbers 1997). The ‘Church Declaration’ addressed churches, religious organisations and communities as the bearers of rights within the religious sphere and acknowledged them as special actors in their own right. At the same time, the national legal regulation of religion was explicitly confirmed and recognised; the member states retain their competence to specify which group counts as a faith community and which legal-organisational form communities should have. The 11th Declaration realises the dictum of ‘unity in plurality’ as the basis of European identity. The particular arrangement between the EU and the religious communities that was manifested in the 11th Declaration established a new government–religion relationship and set the agenda for state–religion relationships in Europe. Additionally, the Declaration changed legal grounds and, for example, facilitated anti-discrimination regulations with regard to religion (Official Journal 1999; ‘Community action programme to combat discrimination (2001–2006)’, EC 2000).
2.2.2. The role of religion in Europe
Beyond the mere policy level, religion still matters for European citizens. Although we do not pretend to summarise the complex, longstanding and contested debate on the role of religion and religiousness, we feel safe to say that – contrary to secularisation theses7 – there is no clear empirical evidence of a uniform trend of secularisation within European countries, neither with regard to an overall pattern of decline in membership in religious organisations (e.g., Iannaccone 1991; Davie 2001; Greeley 2002; Halman and Draulans 2006; Pollack 2008) nor to an overall pattern of decline in the significance of religious values and beliefs on the individual level (e.g., Putnam 2000; Fukuyama 2001; Nelson et al.2001; Frejka and Westhoff 2008; Jagodzinski 2009). Especially after the EU-enlargement in 2004, the European Union must be considered as an exceptional case with regard to its particular and pluralistic mixture of the eastern European Orthodox tradition, the Catholicism of western Europe and the Lutheranism of the north and western European nations (Davie 2001: 264; Spohn 2009).
3. Individual religiosity and Europeanness: Data, operationalisation and models
In order to analyse the changing influence of religion on feelings of Europeanness on the individual level, one wants to have micro-level data that comprise individual attitudes towards the EEC/EU as well as self-reports covering different aspects of religion. In order to allow for a systematic test, these data must be comparable over different countries and should cover at least two points in time before and after the Treaties of Maastricht (1992) and Amsterdam (1997). The requirements are best met by the European Value Studies (EVS) from 1990 and 2000. The EVS is a large-scale, longitudinal, cross-country survey that mirrors what Europeans think about different aspects of everyday life.8 For comparative reasons, we included only those countries for which data were available for both years.9
Focussing on the years 1990 and 2000 enables us to see if attitudes changed on average corresponding to the system transformation in 1992 and 1997. It is not the aim of the analysis to test if the dynamics that we identified on the EEC/EU-level in fact influence individual attitudes which is not possible because the available data are not panel data. However, since the Commission's outspoken policy aim is to strengthen social cohesion through religion, this justifies to testing at least whether on the individual level religion supports feelings of inclusion and social cohesion.
3.1. Operationalisation
We understand social cohesion as a property of societies – in this case: of the EEC/EU – that derive from individual feelings, attitudes and beliefs (e.g., Moody and White 2003). We thereby see it as aggregated individual feelings of integration into this community – as Europeanness. These feelings go hand in hand with the degree of people's trust. Trust results from experiences of social relations and influences how individuals relate themselves to others (Coleman 1990; Gambetta 2001; summarising: Newton 2007). In modern large-scale societies, social trust is influenced also by institutional settings and frameworks. Institutional trust indicates that institutions are perceived as working in the way that they are supposed to and that they are legitimately in place (Putnam 2000; Rothstein 2002; Vobruba 2008). The articulation of institutional trust thereby can be understood as an indicator of feelings of integration into the institutional framework of society; expressed trust in the EEC/EU we thus understand as an indication of a person's feelings of integration into the socio-political and institutional sphere of Europe.
The EVS comprises the following question concerning trust in the in the EEC (1990) and the EU (2000): ‘How much confidence do you have into the European Community (1990)/European Union (2000)?’ This stimulus is vague, and the EVS does not provide further information about the respondents’ understanding of the EEC/EU. We read the question as a general expression of individual feelings of trust towards the political and cultural concept of Europe and thereby, of feelings of Europeanness. We operationalise social cohesion as a societal property by the average of the individual trust in the EEC/EU. We are aware of one question not being able to map complex constructs like Europeanness but as spontaneous reactions towards a stimulus, the answers indicate emotional evaluations of the EEC/EU.10
The country variances of people's confidence in the EEC in 1990 and the EU in 2000 is not high (eta2=0.036/0.068) but significant. Between 1990 and 2000, people in general lost confidence in the European integration process by 11 percentage points.
A comparison of the country-specific odds11 show that Europeans changed from enthusiasm to scepticism: in 1990, the probability of positive attitudes on average was 1.3 times as high as for negative attitudes; in 2000 the probability of positive attitudes became 1.18 times lower. While Slovenia changed from scepticism to euphoria, for France, Germany, The Netherlands, Sweden, Poland, the Czech Republic and Bulgaria the odds reversed. This reversal indicates that people in these countries became much more sceptical over time: in Sweden and Finland, the atmosphere changed during the process of becoming EU members. These findings are very much in line with earlier research that states that Europeans became more critical during the 1990s (McLaren 2002; Eichenberg and Dalton 2007) (see Figure 1).
3.2. Influencing factors
On the individual level, we included sex, age and years of education as socio-demographic variables. Additionally, we control for the individual self-assessment on a left-right scale, individual interest in politics, attitude towards the state's responsibility for social provision, individual satisfaction with the job, general trust, and tolerance level for particular groups. Previous research has shown that these factors might influence the assessment of the EU (Nelson et al.2001; McLaren 2002; Eichenberg and Dalton 2007).12 Although we are not interested in the effects of these variables, we control for them in order to ensure that variances in the dependent variable are not due to their influence and that the country-specific differences in confidence do not depend on their country-specific distributions.
In order to analyse which particular aspects of individual religiousness13 are relevant for social cohesion, we distinguish between three dimensions (Pollack 2008: 171): (a) institutional ties to religious groups, churches, congregations (‘membership’ of a denomination: Catholic, Protestant, Orthodox, other and none; dummy-coded); (b) religious attitudes, feelings, ideas and emotions that a person has in daily life (‘conviction’: salience of religious belief14 ); and (c) practices of institutionalised worshipping and private religious practices (‘practice’: frequency of church attendance, coded on a 6-point scale, and praying outside service, four categories).15 These do not cover whether these ideas are shared by others or if they communicate with ‘official’ understandings. They merely provide an insight into how people understand themselves in relation to religious objects and how intense they feel about them.
Although we explicitly neither test secularisation hypotheses nor changes in religiousness in Europe or country-specific developments, we briefly report some trends that we found in order to frame the following analyses: While the frequency of service attendance in total did not change much over time, it shows significant differences among countries. In Malta, Poland, Romania, Ireland, Portugal and Italy, people reported more frequent attendance, in the Czech Republic, Sweden and France, church attendance was below average in both years. The practice of praying outside service as well changed over time: dramatic increase in praying occurred in Bulgaria and the Czech Republic while decrease was observed in Ireland, Poland, Romania and Malta. The salience of religious beliefs increased slightly over time in all countries, except the United Kingdom, Ireland, Germany and Malta. The data of the EVS 1990 and 2000 showed that in all countries, the majority of self-perceived believers still belonged to a church or congregation. Only in Sweden, more than half of the respondents claimed in 1990 to belong but not to believe (at least not in the traditional Christian canon). In Slovakia, Slovenia and Bulgaria, the number of respondents who claimed to believe and to belong increased between 1990 and 2000. In accordance with other studies (Greeley 2002; Pollack 2008), our analysis shows that religiousness in Europe does not follow a uniform trend of secularisation. There does not even seem to be a clear distinction between eastern and western European countries.
These country-specific differences give raise to the question as to if aspects of the institutional, legal and structural framework on the country-level affects individual interests and understandings. We want to know whether the national economic performance or social spending and unemployment rate (GNI; World Bank Data 1990 and 2000) have any influence on the relationship between religiousness and trust in the EEC/EU because with regard to the research summarised above we expect such factors to trigger Europeanness. We cannot test if the institutional framework on the EEC/EU-level influences individual attitudes towards the European integration project, but it is possible to take country-specific conditions regarding religion into consideration: in accordance with research on the impact of state–religion relationships (Fox and Flores 2009), we consider religion to be a multi-level concept comprising official relationships between states and religious communities (mostly churches), the anchoring of religion in constitutions, the division of fiscal jurisdiction, freedom of worship, or the official policies according to religious symbols. These provide the larger institutional framework for individual attitudes and behaviours. Distinguishing among different dimensions of religiousness on the individual level and on the country-level allows a differentiated perspective on the dynamics of religion in Europe which will be analysed in the next step. Accordingly, we take the relationship between state and churches and the religious composition of the population into consideration. Data are from the State-and-Religion Data Set (Fox 2004). Because Catholicism and Protestantism differ with regard to their ideology, church–state relationship, and the way in which they organise communities, we expect not only the degree of religious homogeneity to matter but also that a Catholic majority influences individual Europeanness differently from a Protestant majority (Nelson et al. 2001; Nelson 2004).16
3.3. Hypothesis testing: The multilevel models
In order to analyse country-level effects, we perform a multilevel mixed-effects logistic regression (using STATA). Multilevel modelling (MML) enables a systematic comparative analysis beyond the comparison of single countries because variables can be defined on different levels and tested simultaneously. Even if country-characteristics are not explicitly tested, MML provides the means to control for differences in country-specific distributions of individual characteristics (Steenbergen and Jones 2002). Depending on our dependent variable a logistic regression is required. If, like in this case, country-level variables are included, multilevel logistic regression performs better than logistic regression because logistic regression tends to underestimate standard errors due to the lack of variability in the macro-level indicators (Hox 2002: 103). A minor problem is that the number (N) of included countries is only 20. We follow Snijders and Boskers (1999) arguing that if N is equal to or greater than 10, a random intercept model is preferable to a fixed model of regression analysis.
Tables 1 and 2 display eb-coefficients (odds ratios) for five/six models in 1990/2000, taking different sets of variables into account. The constant presents the initial odds of confidence; the eb-coefficient indicates the change in odds per one unit change in the predictor variable. Model 1 in each table shows ‘empty models’ in order to test whether there are significant country-level variances. In 1990, the initial probability of positive attitudes towards the EEC was 31% higher than the probability of negative attitudes. In 2000, it was 19.1% times lower. While in 1990, 4.6% of the total variance was due to country-level differences, it was 9% in 2000. Models 2–4 display different sets of individual variables for each year without any country-level controls. In Models 5 and 6, we test country-level variables step-wise to see if they have a significant impact independently of each other. Since we are not interested in the individual-level control variables, we do not comment on them. Their effects are in line with previous research on factors of EU-scepticism (Nelson et al.2001; McLaren 2002; Eichenberg and Dalton 2007). The country-level influences become even more accentuated when entering individual-level variables.
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | |
|---|---|---|---|---|---|
| Gender | 0.899*** | 0.894*** | 0.875*** | 0.874*** | |
| Age | 1.000 | 1.000 | 1.000 | 1.000 | |
| Education | 1.005** | 1.000** | 1.006** | 1.006** | |
| Political interest | 0.888*** | 0.886*** | 0.884*** | 0.883*** | |
| Left/Right self assesment | 1.055*** | 1.054*** | 1.051*** | 1.050*** | |
| State's responibility | 0.971*** | 0.971*** | 0.972*** | 0.972*** | |
| Job satisfaction | 0.998 | 0.998 | 1.000 | 1.000 | |
| Trust in others | 1.223*** | 1.223*** | 1.208*** | 1.207*** | |
| Tolerance of neighbourhood | 0.717*** | 0.714*** | 0.698*** | 0.699*** | |
| Religiousness | 0.999** | ||||
| Catholic | 1.141*** | ||||
| Protestant | 1.026 | ||||
| Orthodox | 1.027 | ||||
| Other | 0.834** | ||||
| Frequence | 1.035*** | 1.036*** | |||
| Believe | 1.212*** | 1.212*** | |||
| Prayer | 1.000 | 1.000 | |||
| % Catholics | 0.940* | ||||
| Constant | 1.308*** | 2.128** | 1.968** | 1.784** | 2.145** |
| Variation country level | 0.162 | 0.197 | 0.180 | 0.186 | 0.156 |
| % of total variation | 4.692 | 5.650 | 5.187 | 5.351 | 4.527 |
| Number of groups | 20 | 20 | 20 | 20 | 20 |
| Wald χ2 of the model | 872.25*** | 948.05*** | 692.1*** | 541.64*** | 684.1*** |
| N | 26450 | 24919 | 24919 | 24836 | 24836 |
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | |
|---|---|---|---|---|---|
| Gender | 0.899*** | 0.894*** | 0.875*** | 0.874*** | |
| Age | 1.000 | 1.000 | 1.000 | 1.000 | |
| Education | 1.005** | 1.000** | 1.006** | 1.006** | |
| Political interest | 0.888*** | 0.886*** | 0.884*** | 0.883*** | |
| Left/Right self assesment | 1.055*** | 1.054*** | 1.051*** | 1.050*** | |
| State's responibility | 0.971*** | 0.971*** | 0.972*** | 0.972*** | |
| Job satisfaction | 0.998 | 0.998 | 1.000 | 1.000 | |
| Trust in others | 1.223*** | 1.223*** | 1.208*** | 1.207*** | |
| Tolerance of neighbourhood | 0.717*** | 0.714*** | 0.698*** | 0.699*** | |
| Religiousness | 0.999** | ||||
| Catholic | 1.141*** | ||||
| Protestant | 1.026 | ||||
| Orthodox | 1.027 | ||||
| Other | 0.834** | ||||
| Frequence | 1.035*** | 1.036*** | |||
| Believe | 1.212*** | 1.212*** | |||
| Prayer | 1.000 | 1.000 | |||
| % Catholics | 0.940* | ||||
| Constant | 1.308*** | 2.128** | 1.968** | 1.784** | 2.145** |
| Variation country level | 0.162 | 0.197 | 0.180 | 0.186 | 0.156 |
| % of total variation | 4.692 | 5.650 | 5.187 | 5.351 | 4.527 |
| Number of groups | 20 | 20 | 20 | 20 | 20 |
| Wald χ2 of the model | 872.25*** | 948.05*** | 692.1*** | 541.64*** | 684.1*** |
| N | 26450 | 24919 | 24919 | 24836 | 24836 |
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | Model 6 | |
|---|---|---|---|---|---|---|
| Gender | 0.956 | 0.946 | 0.930* | 0.929* | 0.928* | |
| Age | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | |
| Education | 1.016** | 1.017** | 1.016** | 1.016** | 1.016** | |
| Political interest | 0.920** | 0.921** | 0.920** | 0.926** | 0.928** | |
| Left/Right self assesment | 1.049** | 1.047** | 1.045** | 1.045** | 1.046** | |
| State's responibility | 0.951** | 0.951** | 0.951** | 0.950** | 0.950** | |
| Job satisfaction | 0.998 | 0.999 | 0.999 | 0.999 | 0.999 | |
| Trust in others | 1.269** | 1.267** | 1.268** | 1.269** | 1.275** | |
| Tolerance of neighbourhood | 0.771** | 0.759** | 0.765** | 0.769** | 0.758** | |
| Religiousness | 0.999 | |||||
| Catholic | 1.203** | |||||
| Protestant | 1.175* | |||||
| Orthodox | 0.775* | |||||
| Other | 0.912 | |||||
| Frequence | 0.999 | 0.999 | 0.999 | |||
| Believe | 1.416* | 1.409** | 1.410* | |||
| Prayer | 0.998 | 0.998 | 0.998 | |||
| % Catholics | 1.014** | |||||
| % Protestants | 0.984** | |||||
| Constant | 0.807 | 0.618 | 0.552 | 0.580 | 0.288 | 0.825 |
| Variation country level | 0.308 | 0.429 | 0.412 | 0.396 | 0.092 | 0.153 |
| % of total variation | 8.565 | 11.547 | 11.138 | 10.733 | 2.719 | 4.455 |
| Wald χ2 of the model | 1458.010 | 1606.85 | 900.66 | 1372.580 | 423.34 | 561.84 |
| N | 22613 | 20927 | 20927 | 20927 | 20927 | 20927 |
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | Model 6 | |
|---|---|---|---|---|---|---|
| Gender | 0.956 | 0.946 | 0.930* | 0.929* | 0.928* | |
| Age | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | |
| Education | 1.016** | 1.017** | 1.016** | 1.016** | 1.016** | |
| Political interest | 0.920** | 0.921** | 0.920** | 0.926** | 0.928** | |
| Left/Right self assesment | 1.049** | 1.047** | 1.045** | 1.045** | 1.046** | |
| State's responibility | 0.951** | 0.951** | 0.951** | 0.950** | 0.950** | |
| Job satisfaction | 0.998 | 0.999 | 0.999 | 0.999 | 0.999 | |
| Trust in others | 1.269** | 1.267** | 1.268** | 1.269** | 1.275** | |
| Tolerance of neighbourhood | 0.771** | 0.759** | 0.765** | 0.769** | 0.758** | |
| Religiousness | 0.999 | |||||
| Catholic | 1.203** | |||||
| Protestant | 1.175* | |||||
| Orthodox | 0.775* | |||||
| Other | 0.912 | |||||
| Frequence | 0.999 | 0.999 | 0.999 | |||
| Believe | 1.416* | 1.409** | 1.410* | |||
| Prayer | 0.998 | 0.998 | 0.998 | |||
| % Catholics | 1.014** | |||||
| % Protestants | 0.984** | |||||
| Constant | 0.807 | 0.618 | 0.552 | 0.580 | 0.288 | 0.825 |
| Variation country level | 0.308 | 0.429 | 0.412 | 0.396 | 0.092 | 0.153 |
| % of total variation | 8.565 | 11.547 | 11.138 | 10.733 | 2.719 | 4.455 |
| Wald χ2 of the model | 1458.010 | 1606.85 | 900.66 | 1372.580 | 423.34 | 561.84 |
| N | 22613 | 20927 | 20927 | 20927 | 20927 | 20927 |
Regarding religion, Models 2 and 3 show that individual religiosity influences EEC/EU-trust: The slightly negative but significant effect of self-perception as a religious person in 1990 became insignificant in 2000; the impact of denomination increased during the same period. Compared to non-believers, being Catholic raised the chances of confidence by 14% in 1990 and almost 21% in 2000. This confirms earlier findings by Nelson et al. (2001) about the role Catholicism played in European integration. In 2000, being Protestant has a positive and significant impact on EU-trust and being an Orthodox reduces the trust-level significantly by 23%. A more differentiated view on the effects of individual religiousness shows that in 1990, both frequency of church attendance and the significance of individual beliefs influenced the trust-level positively (by 3.4 and 21%). In 2000, only the significance of individual beliefs increased the chance of confidence, now by 41% (see Table 3).
| Denomination | 1990 | 2000 | Change in percentage points |
|---|---|---|---|
| Does not apply | 56.7% | 42.0% | −14.7 |
| Catholic | 61.0% | 54.3% | −6.7 |
| Protestant | 50.6% | 33.3% | −17.3 |
| Orthodox | 48.0% | 40.2% | −7.8 |
| Other | 49.1% | 38.3% | −10.8 |
| Denomination | 1990 | 2000 | Change in percentage points |
|---|---|---|---|
| Does not apply | 56.7% | 42.0% | −14.7 |
| Catholic | 61.0% | 54.3% | −6.7 |
| Protestant | 50.6% | 33.3% | −17.3 |
| Orthodox | 48.0% | 40.2% | −7.8 |
| Other | 49.1% | 38.3% | −10.8 |
Including country-level variables does not change the impact of individual religiousness. However, neither a country's economic performance (GNI), the percentage of social spending, the unemployment rate nor the relationship between state and churches has any significant impact.17
The multilevel analysis presented here indicates, however, that the religious composition of the population influenced EU-confidence. Models 5 and 6 show that, while in 1990 a catholic majority decreased Europeanness slightly, in 2000 every additional percentage point of Catholics in a country increased the confidence by 1.4% (Model 5) and every one percentage point of Protestants reduced it by 1.6% (Model 6). While in 1990, the percentage of Catholics expressing trust did not vary systematically among countries, the percentage became higher in more religiously homogeneous Catholic countries in 2000. At the same time, respondents were more sceptical in countries with a higher percentage of Protestants. Further analysis showed that although EU lost credibility over time, the loss was less large among Catholics.18
4. Discussion: Religion matters for the EU-integration process
Our analysis showed that the period between 1990 and 2000 turned out to be a period of different and significant dynamics: while on the EEC/EU-level the integration project turned into a political, social and even cultural project, social cohesion declined between 1990 and 2000. At the same time as the Commission established religion as a means in order to realise integration, religiousness in form of denomination and practices became more important for Europeanness on the individual level.
On the level of legal documents and policy papers, we identified changes in strategy with far-reaching consequences that can be seen as a tendency towards governmental de-secularisation. Starting with the Tindemans Report, the EEC was transformed into a political union during the 1990s. Part of this political augmentation was the ‘discovery of religion’ as a possible foundation of social cohesion. With the initiative ‘A Soul for Europe’ the Commission established a legally rather insignificant but politically important initiative of involving the churches as organisations. This initiative resulted in the ‘Declaration on the status of churches’ in 1997 which became part of community law in the Treaty of Lisbon in 2008. The Declaration marked a two-fold turning point: (a) religious communities became targets of European legal frameworks and were actively involved in European policies and (b) ‘religion’ became a resource that framed European integration in a different way. This frame became even more salient but highly disputed during the political and public quarrels that followed the initiative to establish religious semantics as an integral part of the Preamble of a Constitution for Europe in 2004 (Spohn 2009).
Although we cannot assess if and how the dynamics on the EC-level correspond with individual-level developments, we are safe to say that during the same time, on the individual level of European citizens quite different developments took place: overall confidence in the EU declined after the transformation and, thereby, the degree of social cohesion. This gives raise to the still open question as to whether European citizens just lost confidence in the integration project or whether the attitudinal change indicates a change in the understanding of the EU as not only an economic but also a cultural project in fact mirroring EU policies.
At the same time, denomination became more important, while the frequency of church attendance lost its impact on the feeling of inclusion. This suggests that there is a shift from ‘institutional ties’ to ‘belief’, indicating that religiousness in 2000 strengthens the ‘ideational’ dimension of social cohesion within the EU.19 On the nation-state level, the degree of religious homogeneity gained importance and a higher percentage of Catholics increased EU-confidence while a higher percentage of Protestants lowered it in 2000.
These results support earlier findings that on a country-level, Catholic countries seem to be more in favour of a European federation than the ‘Protestant latecomers’ (Nelson 2004). This does not come as a surprise since the EU started as a mainly Christian-Democratic project on the political level and still is (e.g., Nelson et al.2001; McCrea 2007; Spohn 2009). It was, however, less obvious that in order to develop its integrative effects, individual religiosity seems to need a homogeneous societal framework, as our results suggest; it is neither denomination nor religiousness per se that supports Europeanness. Context matters here: It is the ‘being part of a larger whole’ – preferably in a country with a high proportion of Catholics – that increases the likelihood of Europeanness. While economic differences on the country-level failed to show any influence, the churches, especially the Catholic, seemed to be strong promoters of EU-integration, although a tight state–church relationship failed to show any impact. The results give raise to the necessity to further investigate the relationship between EU, churches, states and EU-citizens: The EC established a particular dialogue with the churches and thereby, circumvent the states and their sovereignty. At the same time, churches are only one respondent from civil society among others like Unions and Women's Organisations.
Although both, the institutionalised dialogue between Commission and churches and positive impact of a more homogeneous catholic environment insinuate that religious organisations, churches and congregations provide important mechanisms creating social cohesion, the analysis at hand cannot sufficiently prove this claim. More research is needed in order to explore how the dialogues with civil society organisations work and which role they in fact play for the integration project. However, this raises the question as to whether the inclusion of civil society organisations is a way to disguise the democracy deficit of EU-policies (e.g., Münch 2001).
Additionally, the mechanism by which churches support feelings of inclusion is still unclear. Research on social capital (e.g., Putnam 2000; Fukuyama 2001; Jagodzinski 2009) or Durkheim's classical study (Durkheim 2010 [1912]) give raise to the expectation that religious organisations provide the necessary norms, meanings and, most important, the opportunity structure and the rites and symbols that facilitate interpersonal attachment.
With regard to social cohesion, the findings suggest that confidence in the EU-integration project is stronger among those who feel similar among the (religiously) like-minded. This finding indicates that cultural aspects have become important for European integration but that there are different dynamics on the EEC/EU-level and on the individual level. While on the EEC/EU-level, religion became salient through the inclusion of religious communities and through legislation, religious membership and matched individual beliefs led to more intense feelings of inclusion on the individual level, especially for Catholics.
This latter development puts the conclusion of successfully integrated religious communities into perspective: it was the Commission's aim to integrate mass organisations, and they succeeded in employing the symbolic capital of these organisations. Although the concrete mechanisms remain unclear, the shift from institutional ties to identification with Christian beliefs and values indicates a change in the way people mentally organise their group identities: Europeanness became related positively to Catholic and negatively to Protestant beliefs and thereby, to feelings of belonging to a larger community of like-minded. This interpretation is supported by the findings that denomination and, at the same time, religious homogeneity on the country-level became more important. However, these results give rise to concern as well: while the integration of organised Christian religiousness seems to support the social cohesion of the EU, it may cause exclusion of all those are not included in this religiousness – whether they be Orthodox or non-Christian denominations or non-believers. As Delanty (2006) pointed out: the boundaries between insiders and outsiders in Europe are dynamic and do not only concern territorial borders but identities as well. The developments that followed the New York terrorist attack in September 2001 give rise to the assumption that the inclusive influence of religion developed an increasingly exclusive side-effect: after 2001, especially Islam became ‘the other’ that turned Europe into a Judeo-Christian enterprise and joined Europeanness with Christianity. This affects minorities within the member states as well as the debates about ‘cultural fit’ of future EU-members. In this regard, the integration of churches and religious communities during the 1990s on behalf of the Commission must be considered as a powerful mechanism of European integration that yet bears the danger of becoming an exclusive tool of the integration of an organised Christian minority in Europe. Further research is needed to evaluate the links among the different levels of individual religiousness, state–church relationships and EU-level religious governance in order to explain how religion is used to construct internal cohesion and external boundaries beyond the crucial period of system transformation between 1990 and 2000.
Footnotes
The EC regards ‘religion’ as a policy field and resource for integration rather than an individual belief-system.
Throughout the article, we use the social cohesion and integration interchangeable. We are aware that both concepts refer to different scientific debates.
Here, we focus on the EC as the only EU-level actor because it is the main executive body of the EU with the right to initiate legislation and supervise its compliance.
Besides the Single European Act, the Charter of Fundamental Social Rights (COM(89)248 final) guarantees the social dimension of the integration project.
Title XIV, Article 130, especially 130b, of the Treaty on European Union (1992) and the Social Cohesion Reports which are presented by the EC every 3 years reporting the progress in economical and social cohesion. The first report was published in 1996.
Simultaneously, the EC published the First Social Cohesion Report (1996) in which the EC proposed to create social cohesion through social dialogue with civil society. The initiative ‘A Soul for Europe’ is one such method.
Without intending to go too deeply into the debate, we assert that different notions of secularisation are discussed: According to Weber's (2006 [1922]) thesis of the ‘enchantment of the world’, secularisation refers to the replacement of religious explanations by scientific ones in modern societies. An alternative notion claims that secularisation refers to the progressive differentiation detaching religion from other functional systems of society (e.g., Luhmann 1977). Secularisation, thirdly, is understand as a lack of religious answers to the stable and constant demand for transcendence (Stark and Bainbridge 1987) or, according to Casanova (1994), as the syndrome of decline of membership, decline of influence of churches and religious communities in the public sphere and as a decline of religious values. Secularisation as well is seen as a trend towards privatisation and individualisation of religious beliefs and practices (e.g., Davie 2001).
The EVS has been criticised for its quality. We cross-checked the size of the denominations from the EVS 2000 with data from Fox (2004) and the ISSP-data on religion from 1998 and found some differences that might be due to differences in data collection procedures. Nevertheless, these differences rarely exceed 5 percentage points. A cross-check for the EVS data from 1990 was not justified since ISSP 1991 and EVS 1990 comprise different countries. Additionally, the results for the individual-level control variables are in line with other research. We therefore consider the EVS as a trustworthy data set.
The countries included: France, UK, Germany, Italy, Spain, Portugal, The Netherlands, Belgium, Denmark, Sweden, Finland, Ireland, Poland, Czech Republic, Slovakia, Hungary, Romania, Bulgaria, Malta, and Slovenia. Although all countries were considered potential members, not all of them were part of the EEC/EU in 1990 or 2000. Additionally, the transformation process of the eastern European countries was not completed in 1990. Empirically, however, neither membership nor being a country in transformation shows any significant impact on the individual confidence.
Research in other areas provides evidence that single-item questions are not necessarily inferior to more complex item batteries (e.g., Nagy 2002). Additionally, our results on the decline of trust during the 1990s are very much in line with research using other data. Therefore, we consider the EVS data as trustworthy.
Odds higher than 1 indicate that the probability of positive attitudes exceed the probability of negative attitudes; odds lower than 1 indicate the opposite.
All items are measured on a 10-point scale and are thereby treated as approximately quantitatively measured variables. A check for multicollinearity indicates correlations not higher than ±0.1.
Depending on the research question, religiousness is operationalised as ‘religious commitment, religious tradition and religious salience’ (Nelson et al.2001: 196), as ‘church membership’ (Jagodzinski 2009) or as ‘belonging to a religious orthodoxy and practicing religious rites on one hand and on the other, experiencing religious feelings, experiences and meanings’ (Davie 1990, 2001).
This is captured by an index out of eight questions on transcendental issues that is standardised between 0 and 1; with 0 indicating no salience and 1 meaning ‘high salience’ (Cronbach's α=0.907).
We do not explicitly refer to the content of different religions or to the functional or substantial dimensions of religion (for such differentiation: Berger 2001).
All country-level data are displayed in the Appendix.
Tables are not reported.
An ANOVA analysis showed that the between-group variance of denominations concerning trust in the EEC/EU is significantly larger than the within-group variance. Additionally, we found that the loss in Europeanness among Catholics between 1990 and 2000 is significantly lower than among members of other denominations. results of the analysis are not reported here.
‘Ideational’ components of social cohesion apply to the identification and affiliation of the members with the group. In contrast, the ‘relational’ dimension refers to the relations between individuals (Moody and White 2003: 104).
Appendix A1.
| % Catholics | % Orthodox | % Protestant | % Other | % Non-believers | % Majority religion1 | Official governmental support1 | GNI capita US dollar 19902 | GNI capita US dollar 20002 | |
|---|---|---|---|---|---|---|---|---|---|
| France | 42.62 | 52.73 | 1.74 | 1.18 | 1.74 | 69.6 | 0 | 19130 | 24450 |
| United Kingdom | 15.01 | 13.79 | 64.10 | 0.20 | 6.90 | 53 | 2 | 15480 | 25480 |
| Germany | 39.91 | 21.51 | 36.01 | 0.39 | 2.17 | 35.72 | 1 | 19500 | 25500 |
| Italy | 17.88 | 81.52 | 0.40 | 0.05 | 0.15 | 84.95 | 1 | 16900 | 20890 |
| Spain | 18.05 | 80.79 | 0.92 | 0.00 | 0.25 | 99 | 2 | 11210 | 15420 |
| Portugal | 11.43 | 85.86 | 0.30 | 0.00 | 2.41 | 97 | 2 | 6430 | 11590 |
| The Netherlands | 55.03 | 22.03 | 17.75 | 0.00 | 5.18 | 34.5 | 0 | 17800 | 26580 |
| Belgium | 35.69 | 55.30 | 3.27 | 0.37 | 5.38 | 80.9 | 1 | 17930 | 25360 |
| Denmark | 10.08 | 0.79 | 87.45 | 0.00 | 1.68 | 86.27 | 4 | 22770 | 31840 |
| Sweden | 25.32 | 1.58 | 71.43 | 0.49 | 1.18 | 84 | 4 | 24590 | 35860 |
| Finland | 11.72 | 0.10 | 85.47 | 1.07 | 1.65 | 89 | 3 | 24890 | 29280 |
| Ireland | 6.93 | 89.01 | 2.48 | 0.20 | 1.39 | 92 | 2 | 23860 | 25470 |
| Poland | 4.57 | 94.06 | 0.73 | 0.27 | 0.37 | 92.2 | 2 | 11400 | 23160 |
| Czech Republic | 64.89 | 29.83 | 4.60 | 0.11 | 0.58 | 40 | 1 | 99 | 4570 |
| Slovakia | 23.07 | 64.16 | 11.42 | 0.83 | 0.53 | 67 | 9 | 99 | 5800 |
| Hungary | 42.51 | 39.10 | 17.99 | 0.20 | 0.20 | 67.5 | 1 | 3170 | 3850 |
| Romania | 2.47 | 7.49 | 3.44 | 85.55 | 1.06 | 77 | 2 | 2720 | 4660 |
| Bulgaria | 29.89 | 0.30 | 0.80 | 60.48 | 8.53 | 82 | 2 | 1620 | 1690 |
| Malta | 1.30 | 97.70 | 1.00 | 0.00 | 0.00 | 91 | 9 | 2140 | 1600 |
| Slovenia | 30.04 | 66.37 | 0.90 | 1.60 | 1.10 | 76 | 9 | 210 | 240 |
| Total | 26.78 | 46.10 | 18.48 | 6.59 | 2.05 | 72.9 | 2.5 | 2140 | 1600 |
| % Catholics | % Orthodox | % Protestant | % Other | % Non-believers | % Majority religion1 | Official governmental support1 | GNI capita US dollar 19902 | GNI capita US dollar 20002 | |
|---|---|---|---|---|---|---|---|---|---|
| France | 42.62 | 52.73 | 1.74 | 1.18 | 1.74 | 69.6 | 0 | 19130 | 24450 |
| United Kingdom | 15.01 | 13.79 | 64.10 | 0.20 | 6.90 | 53 | 2 | 15480 | 25480 |
| Germany | 39.91 | 21.51 | 36.01 | 0.39 | 2.17 | 35.72 | 1 | 19500 | 25500 |
| Italy | 17.88 | 81.52 | 0.40 | 0.05 | 0.15 | 84.95 | 1 | 16900 | 20890 |
| Spain | 18.05 | 80.79 | 0.92 | 0.00 | 0.25 | 99 | 2 | 11210 | 15420 |
| Portugal | 11.43 | 85.86 | 0.30 | 0.00 | 2.41 | 97 | 2 | 6430 | 11590 |
| The Netherlands | 55.03 | 22.03 | 17.75 | 0.00 | 5.18 | 34.5 | 0 | 17800 | 26580 |
| Belgium | 35.69 | 55.30 | 3.27 | 0.37 | 5.38 | 80.9 | 1 | 17930 | 25360 |
| Denmark | 10.08 | 0.79 | 87.45 | 0.00 | 1.68 | 86.27 | 4 | 22770 | 31840 |
| Sweden | 25.32 | 1.58 | 71.43 | 0.49 | 1.18 | 84 | 4 | 24590 | 35860 |
| Finland | 11.72 | 0.10 | 85.47 | 1.07 | 1.65 | 89 | 3 | 24890 | 29280 |
| Ireland | 6.93 | 89.01 | 2.48 | 0.20 | 1.39 | 92 | 2 | 23860 | 25470 |
| Poland | 4.57 | 94.06 | 0.73 | 0.27 | 0.37 | 92.2 | 2 | 11400 | 23160 |
| Czech Republic | 64.89 | 29.83 | 4.60 | 0.11 | 0.58 | 40 | 1 | 99 | 4570 |
| Slovakia | 23.07 | 64.16 | 11.42 | 0.83 | 0.53 | 67 | 9 | 99 | 5800 |
| Hungary | 42.51 | 39.10 | 17.99 | 0.20 | 0.20 | 67.5 | 1 | 3170 | 3850 |
| Romania | 2.47 | 7.49 | 3.44 | 85.55 | 1.06 | 77 | 2 | 2720 | 4660 |
| Bulgaria | 29.89 | 0.30 | 0.80 | 60.48 | 8.53 | 82 | 2 | 1620 | 1690 |
| Malta | 1.30 | 97.70 | 1.00 | 0.00 | 0.00 | 91 | 9 | 2140 | 1600 |
| Slovenia | 30.04 | 66.37 | 0.90 | 1.60 | 1.10 | 76 | 9 | 210 | 240 |
| Total | 26.78 | 46.10 | 18.48 | 6.59 | 2.05 | 72.9 | 2.5 | 2140 | 1600 |
1Data collected for 2000 (Fox 2004) (There are no reliable comparative data for 1990. We assume that the number of ‘belongers’ (Davis 1990) did not change too much between 1990 and 2000. The data are gathered by Fox (2004) on the basis of CIA Fact Book data that refer to official statistics not to self-assessments of interviewees. Differences to survey data are possible from case to case.)
2World Bank.
References
Florian Grötsch is a Ph.D. student at the Bergische Universität Wuppertal. His main area of research is religion in Europe and within the EU. Additionally, he is interested in European Integration and the role and impact that institutions and legal texts play for political organisations.
Annette Schnabel is Professor at the Department of Sociology at the Bergische Universität Wuppertal. Her research interests cover questions of why people join social movements, how they form their national and religious identities and how emotions and rational choices interact. The impact of political and social institutions and their formation form a second focal point of interest. Her research encompasses empirical analyses as well as their impact on social theory.